Leveraging the IT department to reduce operation-caused variation.

One continuing trend in electronics manufacturing services is the increasing role IT-related solutions have in supporting a Lean manufacturing-driven organizational culture. This is particularly true of proprietary solutions that automate processes in ways that minimize normally occurring variation or help eliminate non-value-added activity.

One example of this is SigmaTron International’s proprietary Manufacturing Execution System (MES) system known as Tango, whose Phase III system went live at the EMS company’s Elk Grove Village (IL) facility in June. The overarching goal of Tango is to centralize tools used throughout the company for production management, while adding enough flexibility via customization to address facility-specific or customer-specific situations.

The EMS company uses a combination of proprietary and internally developed systems for enterprise and shop-floor management. All facilities utilize a common ERP system, plus third-party product lifecycle management (PLM) tools. Tango is now deployed as an app accessible through computers and handheld devices. It interfaces with the company’s ERP system, and the goal is to have it be the frontend interface for production personnel for all production-related data collection and tracking activity.

Its Phase III iteration strengthens its ability to support data collection and traceability functions in production and eliminate processing defect opportunities in production activities.

Tango acts as a partner with production operators, tracking each assembly through all processes. On assemblies where serialization needs to be added in different formats at different parts of the process, Tango ensures the right serial number is associated with that assembly. Additionally, if any process steps are skipped, or if an operator attempts to add the wrong serial number, the system notifies the operator of the error. The enforced routing warning feature is a poka-yoke that prevents production operators from making routing errors, while the system collects data on each step of the production process. The data collected support traceability requirements, quality trends tracking and real-time visibility into production status.

The IT department has also created specific standalone applications that help poka-yoke unique requirements. For example, some SMT pick-and-place machines come with software that includes feeder verification, while other placement machines lack this feature. IT wrote a program that verifies component and feeder location against machine programming for machines, which don’t include this feature in their proprietary software. This ensures the same verification process is in place throughout the SMT area. However, because it is only applicable to a subset of the company’s SMT equipment, it isn’t integrated into Tango for company-wide deployment.

As this specialized programming has deployed, lessons have been learned. For example, originally the enforced routing feature wasn’t operator-controllable. When a product was removed from the routing at one step and reintegrated, the system automatically jumped into the next step in the routing. In some returned material authorization (RMA) loop cases, there was a need to repeat a prior step and the system wouldn’t allow the part to be scanned. The enforced routing feature has been modified to enable an override in those cases.

Operator training also has a new approach. When a system is introduced, some employees get nervous about learning to use it. On most production floors, some employees are highly computer-literate, while others don’t use computers often. Regular computer users can usually intuitively figure out next steps if they forget part of the training information, but less frequent users don’t have that advantage. During the Phase III implementation, a full week for training was scheduled, and training was divided into multiple sessions to ensure everyone had time to get comfortable with the system. A strong focus was on explaining “why” the system works as it does, so employees understand how their use of the system helps others in the factory. Training materials are highly visual and make it easy for new users to “walk” through a labeled version of the app that helps them see how each screen relates to the activity they are performing.

Meanwhile, IT staff members provided mentoring during the first week of use. “Superusers” took over the role of mentors for the rest of the team when IT wasn’t available. The result has been that IT support calls have gone down substantially in the month following implementation, and IT resources have been freed up to work on additional programming activities.

Minimizing operation-caused variation via IT-designed poka-yokes will continue to gain in popularity. In this example, the process for designing these poka-yokes has multiple steps. Production personnel identify the need, and a central production representative (normally in facility or production management) is charged with communicating the requirements to IT. An initial requirements definition meeting is held with a corporate IT resource, the production representative and the facility-specific IT resource. A flowchart of the process requiring IT support is provided, and a wireframe of the proposed program solution is developed. Once approved, the programming team creates a beta version for test with the production team. Following the beta test period, improvements are made, and the program is released.

Partnering IT and production teams in creating system-resident poka-yokes reduces operator-driven variation and common process defects. Effectively deploying this strategy requires a well-structured development and testing process, and a training methodology designed to ensure users with different computer competency levels have the support they need to learn the new system.

Tom Rovtar is vice president, IT at SigmaTron International (sigmatronintl.com); tom.rovtar@sigmatronintl.com.